Turn Any Doc into Markdown: FastAPI + MarkItDown + .NET (in Docker)
Автор: Bruno Capuano
Загружено: 2025-10-20
Просмотров: 500
In this video, we build a lightweight MarkItDown Server using FastAPI, Docker, and a .NET client to automatically convert any document (PDF, Word, PowerPoint, etc.) into clean, structured Markdown — perfect for AI pipelines, semantic search, or prompt-ready datasets.
We’ll walk through the entire repo step-by-step: from understanding how Microsoft’s MarkItDown library works to running it inside a Docker container and calling it with a C# console app.
🧩 How it works
The FastAPI server (app.py) receives the binary data of a file, runs it through MarkItDown, and returns Markdown text.
The .NET client (Program.cs) sends files using HttpClient and prints or saves the response.
💡 Resources & Links
🔗 MarkItDownServer Repo (this demo): https://github.com/elbruno/MarkItDown...
📘 Microsoft MarkItDown Library (official repo): https://github.com/microsoft/markitdown
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
